Ultrasensitive Heterogeneous Immunoassay Using Photothermal

We previously proposed a photothermal immunoassay based on highly sensitive detection by a photothermal beam deflection method using colloidal gold as...
3 downloads 0 Views 193KB Size
Anal. Chem. 1996, 68, 3063-3067

Ultrasensitive Heterogeneous Immunoassay Using Photothermal Deflection Spectroscopy. 2. Quantitation of Ultratrace Carcinoembryonic Antigen in Human Sera Hiroko Kimura and Shigetaka Matsuzawa

Department of Forensic Medicine, School of Medicine Juntendo University, 2-1-1 Hongo, Bunkyo, Tokyo 113, Japan Chao-Yi Tu, Takehiko Kitamori, and Tsuguo Sawada* Department of Applied Chemistry, Faculty of Engineering, The University of Tokyo, 7-3-1 Hongo, Bunkyo, Tokyo 113, Japan

We previously proposed a photothermal immunoassay based on highly sensitive detection by a photothermal beam deflection method using colloidal gold as a labeling material. The assay system was significantly improved in terms of background reduction of ultrasensitive laser spectrometry through the newly designed reaction and separation procedure using microspheres for a solid phase. Highly sensitive laser spectrometric analysis is often susceptible to interference by even a small amount of admixtures. In this subsequent study, we improved our photothermal immunoassay so that it possesses further selectivity, matching that of the highly sensitive photothermal detection method. It was proved that our method could be applied to determine ultratrace quantities of a carcinoembryonic antigen (CEA) in cancer patients’ and normal sera. The lower detection limit of CEA was 0.078 ng/mL (9.8 amol, absolute quantity). This is 10 times superior to that of any other immunoassay method and has enough detectability to measure a low level of CEA in healthy individuals. Intraassay coefficient of variation is about 8% at eight different concentrations (n ) 6). We assayed 61 serum samples, and there was a good correlation between this method and previously established ELISA. We obtained clearer demarcation of healthy persons and colon cancer patients in terms of the CEA value than was obtained by ELISA. We believe this is the first report in which the laser-based ultrasensitive immunoassay is proved to be applicable to human sera, although the matrix admixtures causing interference in the assay were quite different in each sample serum. Many researchers have studied and developed a new generation of methodological studies which exceeds the conventional immunoassay methods. For example, laser spectrometric detection, such as fluorometry and turbidimetry, which uses a laser beam as a light source, is capable of an extraordinarily high sensitivity. However, the important and fundamental ideas of new methodological studies have been proved only with experimental standard samples. Due to their high sensitivity, these methods could not be applied to actual biological samples, for it often happens that even a small amount of contaminant substances affects these too-sensitive methods. Therefore, in our previous S0003-2700(95)01196-6 CCC: $12.00

© 1996 American Chemical Society

report,1 we proved the methodological supremacy of our photothermal immunoassay in human serum containing IgE as well as in experimental standard samples. We obtained a good sigmoidal IgE quantitative curve, but this was performed with only one serum. This report demonstrates further effectiveness and applicability of our photothermal immunobinding assay for various specimens derived from human sera. In the field of immunological analysis, immunobinding assay is a widely used technique for the detection of proteins involving Western blot analysis. Proteins on the support are detected by using specific antibodies and labels, such as radioactive isotope,2-4 enzyme,5-8 fluorescent,9,10 colloidal gold,11,12 or others.13,14 The immuno-gold labeling has been the simplest, with no need for further steps of color development and no detection of background staining. However, the sensitivity was less than that of other techniques,15 since colloidal gold possesses a strong optical absorption that is suitable for the laser photothermal method and is optically stable in strong radiation. The photoacoustic spectrum of gold ultrafine particles showed1 an absorption peak at 517 nm, and its molar absorptivity was estimated to be about 104 M cm-2. This absorptivity brings about a high sensitivity by using the photothermal method (optical beam deflection), using a 514.5 nm (1) Tu, C.-Y.; Kitamori, T.; Sawada, T.; Kimura, H.; Matsuzawa, S. Anal. Chem. 1993, 65, 3631-3635. (2) Howe, J. G.; Hershey, J. W. B. J. Biol. Chem. 1981, 256, 12836-12839. (3) Herbrink, P.; Van Bussel, F. J.; Warnaar, S. O. J. Immunol. Methods 1982, 48, 293-298. (4) Jahn, R.; Schiebler, W.; Greengard, P. Proc. Natl. Acad. Sci. U.S.A. 1985, 82, 4137-4141. (5) Hawkes, R.; Niday, E.; Gordon, J. Anal. Biochem. 1982, 119, 143-146. (6) Papas, M. G.; Hajkowski, R.; Hockmeyer, W. T. J. Immunol. Methods 1983, 64, 205-214. (7) Blake, M. S.; Johnston, K. H.; Russell-Jones, G. J.; Gotschlich, E. C. Anal. Biochem. 1984, 136, 175-179. (8) Diamandis, E. P.; Christopoulos, T. K.; Bean, C. C. J. Immunol. Methods 1992, 147, 251-259. (9) Nezlin, R.; Mozes, E. J. Immunol. Methods 1995, 184, 273-276. (10) Ogata, A.; Tagoh, H.; Lee, T.; Kuritani, T.; Takahara, Y.; Shimamura, T.; Ikegami, H.; Kurimoto, M.; Yoshizaki, K.; Kishimoto, T. J. Immunol. Methods 1992, 148, 15-22. (11) Brada, D.; Roth, J. Anal. Biochem. 1984, 142, 79-83. (12) Hsu, Y. H. Anal. Biochem. 1984, 142, 221-225. (13) Heinicke, E.; Kumar, U.; Munoz, D. G. J. Immunol. Methods 1992, 152, 227-236. (14) Spira, G.; Gregor, P. D.; Scharff, M. D. J. Immunol. Methods 1993, 165, 263-268. (15) Kimball, S. R.; Rannels, S. R.; Elensky, M. B.; Jefferson, L. S. J. Immunol. Methods 1988, 106, 217-223.

Analytical Chemistry, Vol. 68, No. 17, September 1, 1996 3063

line of Ar laser excitation. Moreover, gold labeling was not affected for long periods of time by the strong laser beam radiation. Therefore, we adapted gold staining for highly sensitive quantitation by the photothermal beam deflection method. We have already reported16 quantitation of IgE and CEA by photothermal beam deflection measurement of dot assay patterns visualized by an enzymatic technique. The minimum detectable concentration of CEA was 0.8 ng/mL. This was about 10 times superior to that obtained with the conventional densitometric method, but the color of the enzymatic substrate to be assayed was liable to fade upon exposure to strong optical radiation. So, we applied1 the same spectroscopic method to an immunobinding assay using colloidal gold as a labeling material and latex microspheres as a solid phase to measure IgG and IgE in human sera. In the assay, the use of colloidal gold as an antibody labeling material makes the assay highly sensitive because of the low background level, without missing antibodies’ activity. The use of latex microspheres as a proteins support increases the adsorbing area of antigens and the sensitivity of immune reactions. This assay system does not need second antibodies or complicated enhancement steps. This report describes our newly designed sandwich-type assay, which uses capture-antibody-coated latex particles and labeledantibody coated with colloidal gold, and its application to CEA assay in human sera. Our sandwich-type assay is based on the measurement of antibody-coating colloidal gold reacted with the complex of antibody-coated latex and the antigen, using photothermal beam deflection spectroscopy.17 Here, the use of captureantibody-coated latex plays an important part in enhancing the specificity and sensitivity of the assay. The merits of the use of latex as a solid phase are described fully in the Results and Discussion. We show, with evidence, the good applicability of this assay method to various practical samples containing many admixtures and unexpected substances. Carcinoembryonic antigen (CEA) is well known18-22 as a diagonstic tool for long-term monitoring of patients with malignant diseases such as colon cancer, pulmonary cancer, and mammary cancer. The need for reliable diagnosis in the disease resulted in the development of sensitive methods for the detection of CEA. A number of CEA assays have been reported,23-26 and now it is possible to use various commercial CEA assay kits with appropriate sensitivity. In our sandwich-type CEA assay, we tested five monoclonal and two polyclonal antibodies for solid-phase antibody and labeling antibody. Using the best combination of antibodies and solid phase, we assayed a total of 61 serum samples obtained from healthy persons and colon cancer patients. There was a good (16) Matsuzawa, S.; Kimura, H.; Tu, C.-Y.; Kitamori, T.; Sawada, T. J. Immunol. Methods 1993, 161, 59-65. (17) Kitamori, T.; Sawada, T. Spectrochim. Acta Rev. 1991, 14, 275-302. (18) Gold, P.; Freedman, S. O. J. Exp. Med. 1965, 122, 467-481. (19) Fuks, A.; Banjo, C.; Shuster, J.; Freedman, S. O.; Gold, P. Biochem. Biophys. Acta 1975, 417, 123-152. (20) Bo ¨rmer, O. P. Clin. Biochem. 1982, 15, 128-132. (21) Primus, F. J.; Kelley, E. A.; Hansen, H. J.; Goldenberg, D. M. Clin. Chem. 1988, 34, 261-264. (22) Letiagin, V. P.; Vystotskaia, I. V. Vestn Ross. Akad. Med. Nauk. 1995, 4, 10-14. (23) Felder, R. A.; MacMillan, R. H., III; Bruns, D. E. Clin. Chem. 1987, 33, 700-704. (24) Mathieu, F.; Mareschal, J. C.; Collet-Cassart, J. Immunol. Methods 1989, 120, 65-69. (25) Bo ¨rmer, O. P.; Nustad, K., J. Immunol. Methods 1990, 127, 171-178. (26) Fuchs, C.; Krapf, F.; Kern, P.; Hoferichter, S.; Jager, W.; Kalden, J. R. Cancer Immunol. Immunother. 1988, 26, 181-164.

3064

Analytical Chemistry, Vol. 68, No. 17, September 1, 1996

correlation between the results of this method and those previously established with ELISA. EXPERIMENTAL SECTION Antigen and Antibodies. CEA from human colon cancer was commercially obtained from Seikagaku Kogyo Co. and diluted with 0.1% bovine serum albumin (BSA) containing a buffer (1/15 M phosphate buffer, pH 7.4, PB). Seven different anti-CEAs were used. Two polyclonal anti-CEAs (IgG fraction of rabbit antiserum, A115, and affinity-purified goat antiserum, CTS-003) were obtained from Dakopatts and Kyowa Hakko, respectively. Five monoclonal anti-CEAs (four mouse IgG1s, CEJ065, 045-A2604F, 136-A2609C, and 136-A2610D, and one IgG2a, 045-A2605K) were obtained from Boehringer Mannheim Biochemica and BiosPacific. Serum Specimens. Human sera containing CEA were obtained from patients of Juntendo University Hospital. Normal blood samples were collected from our colleagues. Colloidal Gold and Gold Labeling Anti-CEA. The gold particles were prepared according to the method of Frens.27 Bright red suspensions of gold particles, 16-20 nm in diameter, were prepared from tetrachloroauric acid. The absorption peak of the photoacoustic spectrum was at 517 nm.1 The labeling of anti-CEA with gold was performed according to the method of Chakraborty et al.28 All of the anti-CEA antibodies were added to 10 mL of gold particle suspension, and the final antibody concentration was adjusted to 0.05 mg/mL. Latex Particles and Synthetic Porous Membranes. A nitrocellulose membrane, type BA85, pore size 0.45 µm (Schleider & Schuell Co.), and a polycarbonate membrane, Nuclepore, pore size 0.6 µm (Costar Corp.), were used. Latex particles, lot no. HD-08, 0.9 µm in diameter, having a specific gravity of 1.50, were obtained from Takeda Chemical Ind. (Osaka, Japan). Antibody-coated latex particles were prepared as described in a previous article.1 One volume of 1% (w/w) latex particle suspension in PB was coated with an equal volume of anti-CEA (1 mg/mL), and after the washing and the removal of extra antibody, one volume of 0.1% casein containing PB was added for blocking. The antibody-coated latex reagent was kept in 0.1% sodium azide at 4 °C for over 6 months. Methods and Procedures. Our assay schema of sandwichtype immunoassay is shown in Figure 1. It was difficult to optimize all the numerous parameters in the procedures. The details of the determination process may be reported elsewhere, but the optimized conditions are as follows. When the nitrocellulose membrane was used as a solid support, 50 µL of diluted anti-CEA was dotted on to the membrane using the filtration device (Sanplatec Corp. DB-100, Nagoya, Japan), and after the removal of the remaining solution by vacuum, the membrane was washed with 0.1% casein containing PB for blocking of the remaining sites. Then, 50 µL of CEA standard solution or human serum samples was applied, and the membrane was washed again with the same buffer. A 100 µL portion of gold labeling anti-CEA was added to the slot in the device and incubated. All incubation processes took about 40 min. The sample spots that reacted with gold were measured by the photothermal spectroscopy method. When the antibody-coated latex particles were used as a solid (27) Frens, G. Nature Phys. Sci. 1973, 241, 20-22. (28) Chakratborty, U. R.; Black, N.; Brooks, H. G., Jr.; Campbell, C.; Gluck, K.; Harmon, F.; Hollenbeck, L.; Lawler, S.; Levison, S.; Mochnal, D.; O’Leary, P.; Puzio, E. D.; Tien, W.; Venturini, A. Ann. Biol. Clin. 1990, 48, 403408.

Figure 2. Determination curves of CEA assay with polyclonal (b) and monoclonal (0) anti-CEA for gold labeling. For the solid support, the same polyclonal rabbit anti-CEA-coated latex microsphere was used. Figure 1. Schematic illutration of sandwich-type immunobinding assay. (a) A conventional method which uses nitrocellulose membrane as the solid phase. (b) A modified method which uses latex microsphere as the solid phase.

phase, 25 µL of 1% of anti-CEA-coated latex particles and 25 µL of CEA standard solution or specimen serum were mixed in a well of a microtiter plate (Nunc) and incubated for 2 h at room temperature. Next, the CEA antigen-adsorbed latex particles were washed with 0.1% casein containing PB and vigorously mixed with 50 µL of gold labeling anti-CEA. ELISA systems (Dynatech Products) equipment was employed for all washing and centrifugation procedures. After the plate had stood for 2 h, the mixture was washed with 0.1% casein containing PB to remove nonreacted gold particles and resuspended in 75 µL, 50 µL of which was blotted onto a nitrocellulose sheet. In this case, the nitrocellulose membrane was used only as a filter sheet. When a polycarbonate sheet was used as a filter, the mixture of latex and gold suspension was blotted without the washing procedure, because the free gold particles (